A quantum computer would make a conventional supercomputer look like a Palm Pilot.© Photodisk

Two physicists have come up with an idea that could crack the as yet unsolved problem of how to build a quantum computer, a device that would make a conventional supercomputer look like a Palm Pilot.

Hans-Andreas Engel and Daniel Loss of the University of Basel in Switzerland have explained how to make a device called a spin-parity meter.

Such an invention would be analogous to the transistors in microchips: a kind of universal component for building quantum-computer circuitry, says theorist David DiVincenzo of IBM's research labs in Yorktown Heights, New York. The imagined component could allow electronic quantum computers to be made from silicon, just as today's microprocessor chips are.

"With a spin-parity meter in hand, quantum computing could be just around the corner," says physicist José Carlos Egues of the University of São Paulo in Brazil.

But at the moment, Engel and Loss's idea is just that: an idea. They have worked out how a spin-parity meter could be put together in principle, and their theoretical calculations show that it could function as a component of a quantum computer. But they haven't made a real device.

Spin cycle

In an electronic quantum computer, information would be encoded in the magnetic state or 'spins' of electrons: 'up' and 'down' spins would substitute for the ones and zeros of binary code in conventional computers.

Because of the wave nature of quantum particles, pairs of electrons can exist not only as discrete pairs of up and down spins but also as mixtures of such combinations, where each electron can be considered to have an up and down spin simultaneously.

This allows a quantum computer access to many more information-encoding states than a conventional computer, and so it can perform some calculations much more quickly and efficiently.

Last year DiVincenzo and his co-workers theorized about an electronic quantum computer that avoided some of the complications of previous proposals². In their scheme, computing is performed not by feeding data into one end of a circuit and looking at how it emerges at the other end, but by looking at the pattern the data makes as it spreads out through a network of components, just as ripples spread from a pebble dropped in water.

To do this one would need to measure the 'parity' of electron pairs (whether their spins are aligned or opposite) without disturbing them. This is what Engel and Loss have now shown how to do.

Quantum jump

Engel and Loss describe a system in which a pair of electrons is placed in a tiny blob of semiconducting material such as silicon, called a quantum dot. There is a second, empty quantum dot nearby. By applying a magnetic field, it is possible to 'tune' the system such that the energy of neighbouring dots is the same if and only if they contain a pair of electrons with opposite spin. This energy equality allows those electrons with opposing spins to jump between quantum dots. Such jumps can be detected by a highly sensitive meter for measuring electrical charge, and the opposing spins of the electrons can therefore be inferred.

Their scheme, says DiVincenzo, is "a little bit of a mindbender" that could change they way others think about how to make quantum computers. "Experimentalists really pay attention to these people," he says.

Egues says that the "breathtaking" experimental advances that have taken place recently in the relevant electronic technologies "definitely pave the way to an actual implementation" of the idea. 

University of São Paulo, Brazil

• References

  1. Engel H. A., Loss D., et al. Science, 309. 586 - 588 (2005). | Article | PubMed | ChemPort |
  2. Beenakker C. W. J., et al. Phys. Rev. Lett, 93. 020501 (2004). | Article | PubMed | ChemPort |